1. Summary. Eradication of virus-infected cells by cytotoxic T cells involves recognition of Class I MHC products loaded with pathogen-derived peptide cargo. Herpesviruses, in particular human cytomegalovirus (HCMV), display immunoevasive behavior that contributes to their pathogenicity. This behavior relies on the elimination of Class I MHC products from the surface of the infected cell. HCMV accomplishes this through the action of at least two small type I membrane glycoproteins, US2 and US11. Extraction of Class I MHC molecules from the endoplasmic reticulum (ER) occurs under the agency of US2 and US11, followed by delivery to the cytoplasm, where the Class I MHC heavy chains are destroyed by ubiquitin-dependent proteasomal proteolysis. We have identified cellular proteins essential for dislocation of Class I MHC molecules through isolation of proteins that interact with US2 and US11. Their mode of action, their proximity relationships to one another and their exact role in the dislocation process are not known. We shall study these aspects in Aim 1, using immunochemical and genetic approaches, including the generation of conditional alleles in a mouse model, as already accomplished for mouse Derlin-2. We have further developed new biochemical methods that enable the selective tagging, in a genetically specified manner, of proteins of interest with affinity tags and photoactivable crosslinkers. Thus in Aim 2, this method, referred to as sortagging, will be used to explore proximity relationships of proteins involved in dislocation in a manner not hitherto possible. Mechanisms responsible for extraction of proteins from the ER, including those that underlie the unfolded protein response, may be used also for crosspresentation by Class I MHC molecules. In Aim 3 we shall explore this possible interconnection. We shall also examine the role of antibodies in facilitating crosspresentation, using newly developed animal models, including a knock-in mouse model in which all B cells are poised to produce anti-ovalbumin antibodies by VDJ and VJ replacement in the Ig heavy chain and light chain loci, respectively. A second model concerns agammaglobulinemic mice that nonetheless possess a normally diverse set of B lymphocytes. This model allows a distinction between the antigen presenting function of B cells and the contribution of their Ig output to crosspresentation. Finally, in Aim 4 we propose a radically new model for how misfolded proteins leave the ER, based on preliminary data that suggest involvement of US11- and US2-interacting proteins in lipid droplet formation and/or lipid metabolism, respectively. This model would help explain data that have so far resisted a straightforward interpretation within current cell biological dogma. Our model leads to predictions that will be tested by experiment, and so presents the logical complement to experiments conducted under Aims 1 and 2. Combined, the proposed experiments will shed light on aspects of Class I MHC restricted antigen presentation that are poorly understood from a biochemical and cell biological perspective. Project narrative: A molecular description of how cells of the immune system handle foreign antigens is important not only for our understanding of immune responses against pathogens, but also for autoimmunity. The machinery responsible for antigen processing and presentation can be targeted for therapeutic intervention or for improved vaccine development only if we know and understand its component parts. The experiments proposed here will help identify and characterize as yet poorly understood players in antigen presentation, and will thus open up new areas for experimental manipulation of immune responses in health and disease.